Abstract
In this study, a thorough and definitive evaluation of Predictive Optimal Energy Management Strategy (POEMS) applications in connected vehicles using 10 to 20 s predicted velocity is conducted for a Hybrid Electric Vehicle (HEV). The presented methodology includes synchronous datasets gathered in Fort Collins, Colorado using a test vehicle equipped with sensors to measure ego vehicle position and motion and that of surrounding objects as well as receive Infrastructure to Vehicle (I2V) information. These datasets are utilized to compare the effect of different signal categories on prediction fidelity for different prediction horizons within a POEMS framework. Multiple artificial intelligence (AI) and machine learning (ML) algorithms use the collected data to output future vehicle velocity prediction models. The effects of different combinations of signals and different models on prediction fidelity in various prediction windows are explored. All of these combinations are ultimately addressed where the rubber meets the road: fuel economy (FE) enabled from POEMS. FE optimization is performed using Model Predictive Control (MPC) with a Dynamic Programming (DP) optimizer. FE improvements from MPC control at various prediction time horizons are compared to that of full-cycle DP. All FE results are determined using high-fidelity simulations of an Autonomie 2010 Toyota Prius model. The full-cycle DP POEMS provides the theoretical upper limit on fuel economy (FE) improvement achievable with POEMS but is not currently practical for real-world implementation. Perfect prediction MPC (PP-MPC) represents the upper limit of FE improvement practically achievable with POEMS. Real-Prediction MPC (RP-MPC) can provide nearly equivalent FE improvement when used with high-fidelity predictions. Constant-Velocity MPC (CV-MPC) uses a constant speed prediction and serves as a “null” POEMS. Results showed that RP-MPC, enabled by high-fidelity ego future speed prediction, led to significant FE improvement over baseline nearly matching that of PP-MPC.
Highlights
Method which produces some degree of mis-prediction to produce fuel economy (FE) improvements which are somewhere between those produced by the CV-Model Predictive Control (MPC) and Real-Prediction MPC (RP-MPC) methods
The differences in vehicle future velocity prediction Mean Absolute Error (MAE) between the data groups shown in Figure 7 were relatively small where the differences in FE improvement performance based on those predictions shown in Figure 9 were significant
In order to to demonstrate the function of various implementations, the data available to different types of vehicle were classified, an extensive real-world driving dataset was collected which incorporated said data, machine learning (ML) and Artificial Neural Network (ANN) methods were used to predict the ego vehicle future speed using different groups of data, and the best predictions were used in FE simulation to determine the effectiveness of practically implementable
Summary
Previous research in the area of POEMS has focused on select aspects of Research Gap 1 but no comprehensive study has been performed which concerns the use of real-world data and real-time prediction methods in POEMS. In 2019, Refs.[26,32] showed that high-fidelity predictions were possible through the use of deep Long Short-Term-Memory (LSTM) ANNs. in 2020, a thorough analysis of various combinations of real-world data streams and machine learning techniques [26,27,33]. In order to close the gap, this paper outlines a comprehensive system-level study addressing the interactions between groups of available real-world data, velocity prediction methods, and Optimal EMS methods with respect to the overall system output: FE. This paper will further show that the proposed method is implementable on current vehicles with current technology and has the potential to provide significant FE improvements within the HEV fleet if implemented
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
